Active Contour-based Image Segmentation Algorithm

Image segmentation is an improtant part of image processing. The quality of image segmentation directly affects the later image recognition and other applications.In this paper, we introduced the parametric active contour model, the geometric active contour model, the theory of curve evolution and level set method. This paper mainly studied the geometric active contour model and completed the following two parts:In the first place, we research the classical C-V model. Then we point out its disadvantages and improve it. we have added the square of gradient magnitude to energy functional of classical C-V model, and get the improved C-V model. The improved C-V model not only uses the region information but also uses the gradient information of the image. In this case, it has better performance in image with complex backgroud. The experimental results show that the improved C-V model has better performance than the classical C-V model.Secondly, we research and GACV model. We join statistical information in GACV model and get GACV model based on Bayesian criteria. The experimental results show that the GACV model Based on Bayesian criteria can segment color images successfully. In consideration of the characteristics of GACV model and the features of brain tumor CT images, we use the human-computer interaction model based on GACV to segment the brain tumor CT image. The experimental results show that the GACV model based human-computer interaction can segment the brain tumor CT image effectively and accurately.